DE102004007523B4 - Method for determining the switching time from the storage phase to the regeneration phase of a nitrogen oxide storage catalytic converter and for the diagnosis of its storage behavior - Google Patents

Abstract

method for determining the switching time from the storage phase to Regeneration phase of a nitrogen oxide storage catalyst and the Diagnosing its storage behavior, the nitrogen oxide storage catalyst has a nitrogen oxide filling level and in the exhaust tract of a predominantly arranged with a lean air / fuel mixture operated internal combustion engine is and where its degree of filling while the storage phase continuously by integrating the at any time determined per unit of time stored nitrogen oxide mass and the Switching is made on the basis of the achieved degree of filling, characterized in that the after the regeneration remaining filling level of the storage catalyst as start value for the determination of the degree of filling while the next Storage phase is used.

Description

The The present invention relates to exhaust gas purification of predominantly combustion engines operated with a lean air / fuel mixture, so-called lean-burn engines, which include diesel engines and lean-burned Gasoline engines belong. Lean engines have been developed to reduce fuel consumption. In the optimal case, you should save up to 25% on fuel can be.

One Problem of these engines is the removal of the combustion resulting nitrogen oxides from the lean exhaust gas. The of these engines emitted nitrogen oxides exist depending on the operating condition of the engine to 60 to 95% by volume of nitric oxide. Leave the nitrogen oxides due to the high oxygen content of the lean exhaust gas only very difficult to reduce nitrogen and thus render harmless. A possibility To remove the nitrogen oxides is in the purification of the exhaust gases with the help of a nitrogen oxide storage catalyst.

Nitrogen oxide storage catalysts are well known to those skilled in the art. They contain basic oxides, carbonates or hydroxides of the alkali, alkaline earth metals and / or rare earth metals and a catalytic component, mostly platinum, for oxidation of nitrogen monoxide to nitrogen dioxide during normal operation, the is called while lean operation of the engine with a lean air / fuel mixture. While of the lean operation, the generated nitrogen dioxide in the form of Nitrates of the basic components of the storage catalyst bound. As a nitrogen oxide storage catalyst has only a limited storage capacity, it must from Regenerated time to time, that is, the stored nitrogen oxides have to released again and reduced to nitrogen. this happens by short-term operation of the lean-burn engine with a rich air / fuel mixture. Through the necessary, regular regeneration the storage catalyst is the maximum achievable with a lean-burn engine Savings in fuel consumption limited.

The absorbed by a storage catalyst amount of nitrogen oxides is due to the so-called nitrogen oxide degree of filling, or short filling degree, described. The degree of filling is the relationship the actual stored amount of nitrogen oxides to the amount of nitrogen oxides, the maximum in the catalyst at the prevailing exhaust conditions can be stored.

For the with a nitrogen oxide storage catalyst achievable reduction in the Nitric oxide emissions, it is important to timely regeneration before crossing the storage capacity to initiate the storage catalyst. This is usually a Cross-filling ratio set below the storage capacity of the catalyst. Of the respective, current filling level is during the operation of the storage catalyst determined. exceeds the current filling level the limit fill level, the regeneration of the storage catalyst is initiated. Becomes the limit filling level chosen low, such are the remaining emissions of nitrogen oxides low, but increased through the frequent Regeneration of fuel consumption in an undesirable manner. Close to a limit fill level at the storage capacity the proportion of unreachable nitrogen oxides increases. There is also the danger of being on Reason of an inaccurate determination of the current filling level of the specified limit filling level more frequently exceeded which further increases the residual nitrogen oxide emissions.

The nitrogen oxide filling level present during the storage phase is generally determined continuously by integration of the nitrogen oxide mass stored at any time per unit of time. For this purpose, a mathematical model of the storage process is often used. That's how it describes DE 100 36 453 A1 a method for operating a nitrogen oxide storage catalytic converter of an internal combustion engine, wherein in a first phase of operation (storage phase), the nitrogen oxides are stored from the exhaust gas in the storage catalyst and stored in a second phase of operation (regeneration phase) from the storage catalytic converter. The beginning of the second operating phase is determined on the basis of a nitrogen oxide filling level, which is modeled using a nitrogen oxide storage model. In order to be able to determine the beginning and the end of the second operating phase as accurately and reliably as possible, the nitrogen oxide mass flow behind the storage catalytic converter is detected and the NOx storage model is corrected as a function of the detected value.

Next the problem of determining the optimum timing for switching from the storage phase to the regeneration phase consists of a storage catalytic converter the need to decrease its with increasing operating time memory To monitor (aging).

The aging of the storage catalyst consists of a temporary and a permanent Kom component together. The temporary component is a poisoning by the sulfur compounds contained in the exhaust gas. These form with the basic components of the storage catalyst sulfates in competition with the nitrates. The sulfates are much more stable than the nitrates and can not be removed during normal regeneration of the storage catalyst. They are therefore increasingly reducing the storage capacity for nitrates.

The However, sulfate loading of the storage catalyst can be reversed become. This process is called desulfation. For this must the Catalyst at about 650 ° C heated and the exhaust gas to be greased. Typically, the air ratio λ is at Desulfation in the range between 0.7 and 0.98.

at the permanent aging of a storage catalyst is to a thermal injury the components of the storage catalyst. Overheating will cause the storage materials itself and also the catalytically active precious metals sintered and lose on active surface. This process can not reversed become.

When used in a motor vehicle, it is necessary to monitor the aging of the storage catalyst to initiate desulfation if necessary, to switch the vehicle to stoichiometric operation or, if necessary, to provide a catalyst replacement signal. This process is called OBD (On Board Diagnosis). A suitable method for checking the efficiency of a nitrogen oxide storage catalytic converter, which is arranged in the exhaust gas tract of a lean-burn engine is, for example, in DE 198 23 921 A1 described. For this purpose, the current storage capacity of the nitrogen oxide storage catalyst is determined and diagnosed falls below a predetermined minimum capacity, a faulty nitrogen oxide storage catalyst.

The DE 199 51 544 C1 describes a method for controlling the operation of a NOx storage catalyst. According to this document, during the regeneration of a NOx storage catalytic converter, the time at which a lambda probe signal makes a jump is determined. The amount of regenerant supplied up to this jump is a measure of a NOx useful storage capacity. The NOx Nutzspeicherkapazität serves to determine the degree of loading of the storage catalyst during the storage phase using a model calculation.

It Object of the present invention, the accuracy of the Determination of the switching time from the storage phase to the regeneration phase a nitrogen oxide storage catalyst to improve and thus a possibility for the Further optimization of fuel consumption during operation of a Nitrogen storage catalyst while maintaining lower To open up nitrogen oxide emissions. Furthermore the invention should also be a possibility for on-board diagnosis of the performance of the nitrogen oxide storage catalytic converter deliver.

These The object is achieved by an improved method for determining the Switchover time from the storage phase to the regeneration phase a nitrogen oxide storage catalytic converter and to diagnose its storage behavior solved. The nitrogen oxide storage catalyst is here in the exhaust tract of a predominantly with a lean air / fuel mixture operated internal combustion engine arranged and stores during the Storage phase, the nitrogen oxides contained in the exhaust and sets the Nitrogen oxides in the regeneration phase, during which the internal combustion engine for one Regeneration period operated with a rich air / fuel mixture becomes free again and catalytically transforms it. The degree of filling is while the storage phase continuously by integrating the to each Time per unit of time stored nitrogen oxide mass determined and the changeover is made on the basis of the achieved degree of filling.

The Method is characterized in that after the regeneration remaining filling level of the storage catalyst as a starting value for determining the degree of filling while the next Storage phase is used.

The present method is in both gasoline lean burn engines as applicable to diesel engines.

In order to determine the nitrogen oxide filling level F of the storage catalytic converter during the storage phase (lean operation of the engine), the nitrogen oxide mass stored at each time per unit time can be determined from the respective nitrogen oxide conversion U and the nitrogen oxide mass flow in the exhaust gas upstream of the catalytic converter respective sales among other things of the already existing degree of filling, the exhaust gas temperature, the air ratio λ of the exhaust gas and the nitrogen oxide mass flow depends.

The nitrogen oxide conversion U is defined according to equation (1) m _{NOx, before} the nitrogen oxide mass flow before and m _{NOx, behind} the nitrogen oxide mass flow behind the storage catalytic converter:

The nitrogen oxide conversion is maximum when the storage catalyst is completely empty. With increasing filling of the storage catalyst, the conversion decreases to zero, when the degree of filling has assumed the value F = 1. The nitrogen oxides contained in the exhaust gas then pass unhindered through the storage catalytic converter. The change of the nitrogen oxide degree of filling .DELTA.F during a time interval .DELTA.t results according to equation (2) from the nitrogen oxide conversion U by multiplication with the currently prevailing nitrogen oxide mass flow before the catalyst m _{NOx before} . ΔF = U (F, T, λ) · m NOx, before (T) · At. (2)

Of the filling level at the present time can be done through integration over the since the last regeneration operating conditions to be obtained. Upon reaching a limit filling level, the storage phase finished and initiated the regeneration phase.

At the Cross-filling ratio it can be a one-time for the determined storage catalyst determined, fixed value act. However, because the storage capacity of the storage catalyst is not just a material property, but also depends on the currently prevailing operating conditions is the limit filling level preferably in dependence determined by these operating conditions.

at with petrol-powered lean-burn engines is the duration of the storage phase between about 60 and 120 seconds, during the duration of the regeneration phase about 1 to 5 seconds. While the storage phase becomes the engine with a lean air / fuel mixture with an air ratio λ of usually over 1.3 operated. The air ratio here is stoichiometric conditions normalized air / fuel ratio. to Regeneration of the catalyst becomes the engine with a rich air / fuel mixture operated, that is the air ratio of the exhaust gas leaving the engine is smaller 1. Usually a value between 0.7 and 0.98 is selected. The working temperature typical storage catalysts is between 150 and 550 ° C.

The Regeneration of the storage catalyst can be operated with gasoline Lean-burn engines can be carried out without problems, as this engine type be operated stably even with a rich air / fuel mixture can. The regeneration of the storage catalyst is usually Completely, this means the previously stored nitrogen oxides become almost during regeneration Completely released again.

The nitrogen oxide content of the exhaust gas of diesel engines is significantly lower than that of gasoline lean-burn engines. Accordingly, the storage phase for the storage catalytic converter in the exhaust tract of a diesel engine with the same storage capacity can be up to 300 seconds or more. Since a diesel engine is not very stable when operating with a rich air / fuel mixture, the regeneration of the storage catalyst is difficult. Here, it often happens that the regeneration must be aborted prematurely due to changes in the operating state of the engine, for example, and the storage catalytic converter at the beginning of the new storage phase therefore still has a residual F F _residual nitrogen oxides, according to the invention as the starting value for determining the Nitrogen oxide filling level is used during the next storage phase.

The consideration the remaining degree of filling after regeneration as starting value for the determination of the nitrogen oxide filling level while the next Storage phase leads in the case of petrol-powered lean-burn engines, only minor changes, because the storage catalyst in these engines during regeneration in the Always almost complete can be emptied. In the case of diesel engines is the complete emptying of the storage catalyst during However, regeneration is not always guaranteed. Here leads the inventive consideration the remaining degree of filling in determining the switching time to a much better Performance of the emission control system and usually also to a reduced fuel consumption.

The remaining degree of filling after regeneration of the storage catalyst can be determined by using a previously determined dependence of the degree of filling on the degree of filling at the beginning of the regeneration of the Re generation period, the air ratio of the exhaust gas during regeneration and the exhaust gas temperature are determined. For this purpose, the previously determined dependency can be stored in the form of empirically determined characteristics fields or as a mathematical model in the electronic control of the engine.

According to equation (2) is used to determine the degree of filling while the storage phase stored at each time per unit time Nitric oxide mass from the respective nitrogen oxide conversion and the nitrogen oxide mass flow determined in the exhaust gas before the catalyst. For this purpose, a Nitrogen sensor located behind the storage catalytic converter in the exhaust system be detected, the nitrogen oxide mass flow behind the catalyst. Of the is necessary to calculate the sales necessary nitrogen oxide mass flow upstream of the catalyst usually in the electronic control of the engine for everyone Operating point in the form of a mathematical model or empirical ascertained characteristics fields and therefore must not necessarily with the help of a second nitric oxide sensor before Catalyst are measured. The emitted nitrogen oxide mass flow the engine's main operating parameters include the Speed, torque (load), the air / fuel ratio supplied to the engine, temperature the intake air, ignition angle, Exhaust gas recirculation, etc.

alternative for metrological Detection of the nitrogen oxide mass flow behind the catalyst and thus also of the nitrogen oxide conversion can these sizes too be determined purely mathematically. For this purpose, the dependency must first the turnover of the degree of filling of the storage catalyst for different operating states be determined of the engine. The storage catalyst is first conditioned and then the conversion starting from a completely emptied storage catalyst dependent on of self-adjusting degree of filling determined experimentally. This functional dependence becomes empirical Data, e.g. as characteristic curves, or as a mathematical model deposited in the electronic control of the engine.

The degree of filling of the storage catalytic converter is then calculated during operation during a storage phase by integration of equation (2) over the past operating conditions of the engine since the last regeneration, according to the invention as a boundary condition of the remaining degree of filling at the beginning of the storage phase is taken into account. From the characteristic curves or the mathematical model, the nitrogen oxide mass flow m _Nox belonging to the instantaneous engine operating point is taken for this purpose and the conversion U for the degree of filling already achieved.

The Just described determination of the degree of filling is without recourse possible on measured sales values. However, there is a risk that the calculations are due to the aging of the storage catalyst increasingly from the real conditions remove. Therefore, it is recommended that the nitrogen oxide sales in addition the above-described measurement of the nitrogen oxide mass flow behind the catalyst.

The additional Measurement of the nitrogen oxide mass flow behind the storage catalytic converter, or the measurement of the nitrogen oxide conversion, made possible by Comparison with that from the mathematical model or the empirical Characteristic fields determined nitric oxide sales a diagnosis of storage capacity of the catalyst. A reduction in the storage capacity of the catalyst occurs when the measured nitrogen oxide mass flow is the calculated one Nitrogen oxide mass flow over exceeds a specified period and by a predetermined amount. If this is the case, then first tried the storage capacity by a sulfur regeneration (desulfation) restore. After repeated, unsuccessful sulfur regeneration, a signal can appear be set to replace the storage catalyst. Until the Replacing the catalyst, the engine can be switched to stoichiometric operation be, as experience has shown a damaged storage catalyst still has sufficient catalytic activity for stoichiometric operation.

The previously described method for computational determination of the nitrogen oxide conversion from the characteristic fields or with the help of a mathematical model goes away from that the Relationship between nitric oxide conversion and nitrogen oxide filling level independent of after a regeneration still remaining in the storage catalytic converter filling level is. However, experimental studies by the inventors have shown that the Acceptance of such, from the immediate history of the catalyst independent Relation represents a rough simplification. It was found, that directly after a partial, or incomplete, regeneration of nitric oxide sales of the storage catalyst higher is as according to the dependency the nitrogen oxide conversion of the degree of filling after complete Regeneration would be expected. Only after a certain time approaching his behavior returns to behavior after complete regeneration at. This fact leads to corresponding errors in the integration of the degree of filling and thus to a faulty determination of the switching time from the storage phase to the regeneration phase.

Therefore, a further improvement in the determination of the switching time can be achieved if the respective nitrogen oxide conversion during the storage phase for the present operating condition and time after a previous incomplete regeneration of the storage catalyst according to equation (3) is provided with a correction, the correction K is determined from a previously determined dependence between nitrogen oxide conversion and nitrogen oxide degree of filling after incomplete regeneration ΔF = [U (F, T, λ) + K (F, F rest , T, λ)] · m NOx (t) · Δt, (3) where F _{remainder is} the remaining degree of filling after partial regeneration.

As already explained, it has been found that the correction K decreases as the degree of filling F increases, that is to say as the distance between the instantaneous filling level and the remainder of the filling level F _remains after the last regeneration (F-F _remainder ), so that the behavior of the catalyst increases As the filling level approaches, the behavior of equation (1) approaches again. Therefore, the correction can be approximated as linear or exponential with (F - F _residual ) decreasing function.

Also using equation (3) to integrate the degree of filling over the past operating conditions of the engine, the process can again by measuring the nitrogen oxide mass flow added behind the catalyst become a diagnostic option for the memory to create the catalyst.

The The following examples and the figures serve to better understand the present invention. Show it:

1 : Dependence of the nitrogen oxide conversion on the fill level of the storage catalyst, calculated as NO ₂ in grams per liter of catalyst volume [g / l], for a storage catalyst after complete regeneration

2 : Nitrogen oxide slip [ppm by volume] after storage catalyst during repeated storage cycles with incomplete regeneration

3 : Remaining degree of filling (residual filling level F _residual ) of a storage catalytic converter after incomplete regeneration (partial regeneration) as a function of the degree of filling at the beginning of the regeneration

4 : Dependence of the nitrogen oxide conversion of the already existing degree of filling of a storage catalyst after incomplete regeneration of the same

example

A conventional storage catalyst was prepared by coating a cordierite honeycomb body (62 cells / cm ² equivalent to 400 cpsi) with a barium oxide-based catalyst composition containing catalytically active components of platinum and rhodium on an active alumina.

For the following investigations in a model gas plant, a core of this 2.54 cm diameter, 7.62 cm long catalyst was used. In the model gas plant, the catalyst was subjected at a gas temperature of 250 ° C and a space velocity of 77000 h ^-1 different working cycles of storage and regeneration phases. The gas compositions during these phases can be found in the following table. The catalyst had a storage capacity of about 2.5 grams of NO ₂ per liter of catalyst volume under these exhaust conditions.

Table: Gas composition during storage and regeneration phase

Nitric oxide conversion depending on from the degree of filling after complete Regeneration of the storage catalytic converter

to Determination of the nitrogen oxide conversion depending on the already stored nitrogen oxide mass after each complete regeneration before each measurement, the catalyst became at a gas temperature of 550 ° C conditioned with 15 fat / lean cycles. Thereafter, the nitrogen oxide conversion was during a Storage phase from the nitrogen oxide concentration in front of and behind the Catalyst determined and from the stored on the catalyst Nitrogen oxide mass calculated by integration over the storage phase. The standardization of this nitrogen oxide mass to those at these exhaust conditions available storage capacity of about 2.5 g of NOx per liter of catalyst volume gives the degree of filling of Storage catalyst.

1 shows the corresponding behavior of the catalyst. The nitrogen oxide conversion is nearly 1 (100%) for the fully regenerated catalyst. With increasing occupancy by nitrogen oxides, the nitrogen oxide conversion decreases and approaches 100% as the degree of filling approaches zero.

These Form of functional dependence between nitrogen oxide conversion and degree of filling is almost independent from the nitrogen oxide concentration contained in the exhaust gas and from its Space velocity.

Nitrogen oxide slip after storage catalyst during repeated memory cycles with incomplete regeneration

2 shows the nitrogen oxide slip after catalyst for consecutive cycles of each of a storage phase of 320 seconds duration and a regeneration phase of only 2 seconds.

It is easily recognizable that during a Storage phase of nitrogen oxide slip with increasing occupancy of the Storage catalyst increases. Without regeneration would the Slip finally equal to the nitrogen oxide concentration before the catalyst. The memory of the catalyst would be exhausted.

The However, after every 320 seconds, regeneration is sufficient not enough to completely regenerate the catalyst, that is, the slip at the beginning of the next Storage phase already has a finite value, which increases with increasing Number of performed Work cycles increases, but after a certain number of memory cycles assumes a constant value. In this steady state is while a storage phase stored just as much nitric oxide as during the Regeneration phase can be released again.

The parameters used here (duration of storage and regeneration phase) are typical for use in a diesel vehicle, if the regeneration phase can not be completely completed due to necessary load changes. In this case, a considerable part of the nitrogen oxides stored in the last storage phase remains on the storage catalytic converter. According to the invention, this fact is there borne by the fact that in the pursuit of the next storage phase of the remaining remaining degree of filling of the storage catalyst F _{residual is} taken into account as a starting value in the calculation of the subsequent nitrogen oxide storage.

After incomplete regeneration on the storage catalyst remaining residual level

3 shows the dependence of the still remaining on the storage catalyst in incomplete regeneration residual filling level of the existing on the catalyst at the beginning of the regeneration filling rate derived from such measurements. There is an approximately linear dependence of these two quantities on each other.

To determine the in 3 Dependence shown were two series of duty cycles according to 2 added. In the first series, the storage phase only had a duration of 70 seconds with a 2 second regeneration phase. The measurement points obtained from this are in 3 marked by squares. The storage phase of the second series lasted in each case 320 seconds with constant regeneration duration of only 2 seconds. The corresponding measuring points are indicated by triangles.

Dependence of the nitrogen oxide conversion from the degree of filling the storage catalytic converter after completed, incomplete regeneration of the storage catalyst

The dependence of the nitrogen oxide conversion of the degree of filling of the storage catalyst after each complete regeneration has already been in 1 shown. This dependence is used to calculate the degree of filling during a storage phase by integration over the duration of the storage phase. If a specified limit is reached or exceeded, the regeneration of the storage catalytic converter can be initiated.

The curve of 1 can be used according to the invention approximately for the calculation of the degree of filling after incomplete regeneration. However, more detailed studies by the inventors have shown that directly after a regeneration of the nitrogen oxide conversion is greater than would be expected according to the Restfüllgrades of the catalyst. Therefore, the curve of 1 also determined in the event of incomplete regeneration. For this purpose, a series of work cycles with incomplete regeneration was again recorded and determined from the measured values, the dependence of the nitrogen oxide conversion of the degree of filling.

The curves thus determined shows 4 for the combination of a storage phase of 320 seconds with a regeneration phase of 2 seconds duration compared to the corresponding curve after complete regeneration according to 1 , As this dependency shows, the measured nitrogen oxide conversion directly after incomplete regeneration is significantly greater than would be expected according to the degree of filling after complete regeneration ( 1 ). The increase in nitrogen oxide conversion, however, decreases rapidly with increasing filling level, so that after a short time the nitrogen oxide conversion is equal to the value of complete regeneration.

Therefore, according to the invention after completed incomplete Regeneration during the next storage phase to be calculated storage load by taking into account the described elevation of Nitrogen oxide sales, for example, by a correction element significantly be improved.

Claims (10)

Method for determining the switching time from the storage phase to the regeneration phase of a nitrogen oxide storage catalytic converter and for the diagnosis of its storage behavior, the nitrogen oxide storage catalytic converter having a nitrogen oxide filling level and being arranged in the exhaust gas tract of an internal combustion engine operated predominantly with a lean air / fuel mixture and being Filling degree during the storage phase continuously determined by integration of the stored at any time per unit time nitric oxide mass and the switch is made on the basis of the achieved degree of filling, characterized in that the remaining after the regeneration degree of filling of the storage catalyst as the starting value for determining the degree of filling during the next storage phase is used. Method according to claim 1, characterized in that that the after regeneration still remaining degree of filling of the storage catalyst using a previously determined dependence of the degree of filling from the degree of filling at the beginning of the regeneration, the regeneration period, the air ratio of the exhaust gas during the regeneration and the exhaust gas temperature is determined. Method according to claim 2, characterized in that that the nitrogen oxide mass stored at any time per unit of time from the respective nitrogen oxide conversion and the nitrogen oxide mass flow is determined in the exhaust gas before the catalyst. Method according to claim 3, characterized that to Determination of the nitrogen oxide conversion of the nitrogen oxide mass flow behind measured catalyst and the nitrogen oxide mass flow upstream of the catalyst on the basis of a mathematical model or with the help of a previously empirically determined nitrogen oxide mass flow for the respective operating point is determined. Method according to claim 3, characterized that the Nitric oxide sales for the instantaneous operating point of the motor by means of a mathematical Model or empirical data for a complete one Regenerated storage catalyst determined and from the nitrogen oxide filling level of Storage catalytic converter through integration over since the last regeneration past operating conditions of the engine is calculated. Method according to claim 5, characterized in that that to Determination of the degree of filling the respective nitrogen oxide conversion is provided with a correction, the correction being from a previously determined dependency between nitrogen oxide conversion and degree of filling after incomplete Regeneration is determined. Method according to claim 5 or 6, characterized that in addition the Nitrogen oxide mass flow is measured behind the catalyst. Method according to claim 7, characterized in that that to Diagnosis of storage capacity of the nitrogen oxide storage catalyst, the measured nitrogen oxide mass flow behind the catalyst with a calculated nitrogen oxide mass flow wherein the calculated nitrogen oxide mass flow from that according to claim 5 calculated nitrogen oxide conversion and the nitrogen oxide mass flow before the catalyst is determined. Method according to claim 8, characterized in that that, if the measured nitrogen oxide mass flow the calculated nitrogen oxide mass flow over a period and by a predetermined amount, closed on a reduction in the storage capacity of the catalyst and a sulfur regeneration of the catalyst is triggered. Method according to claim 9, characterized that after repeated, unsuccessful sulfur regeneration of the storage catalyst is exchanged.